Insights into the Mechanisms by Which Clostridial Neurotoxins

Subscriber access provided by University of Colorado Boulder

Article

Insights into the mechanisms by which clostridial neurotoxins discriminate between gangliosides. Joshua R. Burns, Gregory S. Lambert, and Michael R. Baldwin Biochemistry, Just Accepted Manuscript • Publication Date (Web): 25 Apr 2017 Downloaded from http://pubs.acs.org on April 26, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Biochemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Insights into the mechanisms by which clostridial neurotoxins discriminate between gangliosides. Joshua R. Burns # †, Gregory S. Lambert # and Michael R. Baldwin *

Department of Molecular Microbiology and Immunology, University of Missouri School of Medicine, Columbia, Missouri, U.S.A.

ACS Paragon Plus Environment

1

Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 43

ABSTRACT

Botulinum neurotoxins (BoNTs) and Tetanus neurotoxin (TeNT) are the causative agents of the paralytic diseases botulism and tetanus, respectively. Entry of toxins into neurons is mediated through initial interactions with gangliosides, followed by binding to a protein co-receptor. Herein we aimed to understand the mechanism through which individual neurotoxins recognize the carbohydrate motif of gangliosides. Using cell-based and in vitro binding assays, in conjunction with structure-driven site directed mutagenesis, a conserved hydrophobic residue within the BoNTs that contributes to both affinity and specificity towards Sia5-containing gangliosides was identified. We demonstrate that targeted mutations within the Sia5 binding pocket result in the generation of neurotoxins that either bind and enter cells more efficiently (BoNT/A1 and BoNT/B) or display altered ganglioside binding specificity (TeNT). These data support a model in which recognition of Sia5 is largely driven by hydrophobic interactions between the sugar and the Sia5 binding site.


2

Page 3 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

INTRODUCTION Several species of Clostridium, most notably C. botulinum and C. tetani, are renowned for their ability to produce potent neurotoxins

1-3

. There are seven major serotypes of botulinum neurotoxin (BoNT,

types A–G), and a significant number of amino acid variants identified by nucleotide sequencing (commonly termed subtypes)

4-7

. Together with tetanus neurotoxin (TeNT), they form the clostridial

neurotoxin (CNT) family of protein exotoxins. Exposure to botulinum or tetanus neurotoxins results in the life-threatening neuroparalytic diseases of botulism and tetanus, respectively. In their active form, CNTs are composed of an ~100 kDa heavy chain (HC), which includes translocation (HCT) and receptor binding (HCR) domains, linked by a single disulfide bond to an ~50 kDa light chain (LC) that constitutes the catalytic domain 8. Intoxication of neurons is a multistep process involving receptormediated endocytosis, translocation of LC into the cytosol, and proteolytic cleavage of neuronal SNARE (soluble N-ethylmaleimide-sensitive fusion protein attachment protein receptor) proteins. SNARE proteolysis abrogates synaptic vesicle exocytosis, ultimately resulting in the paralysis of the innervated muscle or gland 9-11. Currently, eight subtypes of BoNT/A (termed A1-A8) have been identified, ranging in amino acid similarity from 98% (A1 versus A5) to 84% (A1 versus A3). To date, only a limited number of studies have been performed to compare functional differences between the subtype toxins

12-15

. Analysis of

BoNT subtypes A1-A5 has demonstrated BoNT/A4 to be ~1000-fold less active than A1; BoNT/A3 has a dramatically shorter duration of action within neurons; and BoNT/A2 enters cells more rapidly than A1. Moreover, multiple studies using ex vivo cell cultures, animal models, and human subjects reveal that BoNT/A2 has greater clinical efficacy and less spread of its action to a neighboring muscle as compared with BoNT/A1

16-22

. However, despite the accumulating evidence that subtypes have unique

pharmacologic characteristics, the mechanistic basis for such differences have not yet been defined.


3

Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 43

Neuronal binding of CNTs associated with human disease (BoNT types A, B, E, F and TeNT) is mediated by the HCR domain, which simultaneously recognizes host synaptic vesicle proteins and complex gangliosides present on the presynaptic membrane

10

. Studies performed by multiple

investigators revealed that gangliosides bind to these toxins through a shallow cleft formed by a series of residues referred to as the conserved ganglioside binding motif (GBM; E/D…H/K…SXWY…G)

23-28

.

Although residues within the GBM are necessary for interaction with the GalNAc3-Gal4 moiety of gangliosides, the GBM alone does not account for the high affinity and specificity of ganglioside binding between individual CNTs 25. TeNT stabilizes interaction with ganglioside through formation of an additional hydrogen bond between Asn1219 and the GalNAc3 sugar 24, 25. As a consequence, binding of TeNT to gangliosides containing the core GalNAc3-Gal4 moiety is not influenced by the addition of one or more N-acetylneuraminic acid residues. In contrast, BoNTs associated with human disease (BoNT types A, B, E, and F) bind with high affinity only to gangliosides containing a α2, 3-linked Nacetylneuraminic acid sugar (denoted as Sia5) attached to Gal4

25

. In the case of BoNT/F, binding of

Sia5 is mediated by hydrogen bonds formed to Arg1111 and Arg1256 in addition to hydrophobic interactions with the Sia5 binding site. While Arg1111 and Arg1256 are not conserved, structural data suggest that BoNTs A and B contain alternate residues which perform an analogous function (Figure S1) 25, 27, 28. In the present study, we focused on understanding the mechanism(s) by which individual clostridial neurotoxins bind to, and discriminate between, individual gangliosides. Biochemical characterization of wild-type and mutated HCR domains suggest that hydrogen bonds formed between the toxin and Sia5 make only moderate contributions to the affinity and specificity of the interaction. Rather, a semiconserved hydrophobic residue within the BoNTs was identified that plays an important role in Sia5 recognition. Replacement of isoleucine 1239 in HCR/B with the corresponding phenylalanine residue of BoNTs A, E and F increased affinity for gangliosides GT1b and GD1a by an order of magnitude as


4

Page 5 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

compared to wild-type. Moreover, HCR/BI1239F was shown to both enter cortical neurons and inhibit the activity of BoNT/B holotoxin more efficiently than wild-type HCR/B. The importance of the semiconserved hydrophobic residue was further substantiated by the construction of a novel TeNT variant that preferentially bound gangliosides containing the Sia5 sugar. These data support a model in which recognition of Sia5 by BoNTs is driven primarily through hydrophobic interactions between the protein and the sugar moiety.

MATERIALS AND METHODS Site-directed mutagenesis of BoNT and TeNT HCR domains. A previously described, modified pET28a expression vector (EMD Millipore, Billerica, MA, USA)—hereafter referred to as pET283×FLAG—containing a 3×FLAG epitope immediately downstream of the hexahistidine epitope tag was used as the expression vector for all HCR constructs

29

. E. coli optimized HCR/A1 (aa 870-1295),

HCR/A2 (aa 870-1295), HCR/B (aa 859-1290), HCR/E (aa 849-1251), HCR/F (aa 868-1278), and HCR/T (aa 865-1315) DNA was synthesized (Genscript, Piscataway, NJ, USA) and subcloned into pET28-3×FLAG via unique 5’ KpnI and 3’ XhoI restriction sites. Point mutations were introduced into the BoNT and TeNT HCR sequences by site-directed mutagenesis using the Quikchange® II SiteDirected Mutagenesis Kit (Agilent Technologies, Santa Clara, CA, USA) and confirmed by automated DNA sequencing. In contrast to BoNTs A, B, E, and F, TeNT contains a second binding site that binds gangliosides through terminal sialic acid residues

24, 30, 31

. Therefore, to compare ganglioside binding

mediated solely by the conserved ganglioside binding site (formed by the GBM), all experiments were performed on the HCR/TR1226L background 30. Generation of a chimeric TeNT – BoNT/A protein using Splicing by Overlap Extension-PCR. A chimeric HCR/T protein containing a short sequence of amino acids derived from BoNT/A (1270-1279) ACS Paragon Plus Environment

5

Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 43

was generated by Splicing by Overlap Extension-PCR. For each PCR fragment, primers were constructed such that a 5’ overhang of 30 nucleotides encoding BoNT amino acids was complementary to the 5’ end of the second PCR fragment. In the first round of PCR, two fragments of approximately 1300 and 300 base pairs were amplified and subsequently gel purified. Once purified, a second round of PCR was carried out by mixing equimolar amounts of the two purified products with only primers from the two far ends. The overlapping complementary sequences introduced in round 1 served as primers and the two sequences containing the BoNT insertion were joined. DNA was digested using KpnI and PstI restriction enzymes, ligated into pET28-3×FLAG and verified by automated DNA sequencing of the entire HCR coding region. Expression and purification of HCR proteins. For purification of HCR, E. coli BL-21(DE3) harboring pET28-3×FLAG HCR was inoculated into 2 l of LB supplemented with kanamycin for 3 hours at 37ºC and 250 r.p.m. (an OD600 of ~0.6). Isopropyl β-D-1-thiogalactopyranoside (IPTG) was added to a final concentration of 1 mM and incubation continued overnight following temperature reduction to 16ºC. Bacterial cells were harvested by centrifugation at 6000 ×g for 20 min at 4ºC, lysed by French press, and clarified by centrifugation at 15000 ×g for 30 min at 4ºC. Lysates were filtered through a 0.22 µm cellulose acetate syringe type filter prior to sequential column chromatography: Ni2+nitrilotriacetic acid (NTA) agarose (Qiagen, Germantown, MD, USA) affinity chromatography followed by size exclusion chromatography using a S200-HR gel filtration column (GE Healthcare Bio-Sciences, Pittsburgh, PA, USA). Column fractions containing purified 3×FLAG-HCR were pooled, concentrated using an Amicon type centrifugal filtration device, and dialyzed against 30 mM HEPES-NaOH, 150 mM NaCl, pH 7.6. Purified 3×FLAG-HCRs were stored undiluted at -80ºC until use. A representative SDSPAGE gel showing the relative purity of the HCR domains used in these studies is shown in Figure S2. Circular Dichroism. An AVIV model 202 far-UV spectrometer was used to collect spectra (198–250 nm) of wild-type and mutated HCR domains (0.5 mg/ml) in 10 mM potassium phosphate buffer, pH 7.2. ACS Paragon Plus Environment

6

Page 7 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Spectra were collected at 25°C, using a 1-mm path length quartz cuvette. Baselines were recorded using potassium phosphate buffer and subtracted from the sample spectrum. Measurements were only made down to wavelengths where the instrument dynode voltage indicated the detector was still in its linear range. Background corrected spectra for wild-type and mutated HCR domains are shown in Figure S3.

Ganglioside binding assay. A solid phase ganglioside binding assay was carried out as described previously

13

. Briefly, 5 µg of purified bovine brain gangliosides (Matreya, LLC, State College, PA,

USA) were coated onto a non-protein binding 96 well plate (Corning Costar #3591). Nonspecific binding sites were blocked by incubation with 2% (w/v) bovine serum albumin (BSA) in carbonate buffer solution, pH 9.6. Various concentrations of HCRs were added to the plates and incubated at 4ºC for 90 minutes. Post incubation, 96 well plates were washed three times after which an antibody solution containing mouse anti-FLAG M2 monoclonal antibody (1:8000, Sigma-Aldrich, St. Louis, MO, USA) and goat anti-mouse HRP conjugate (1:10000, Thermo Fisher Scientific, Waltham, MA, USA) was applied and incubated for 20 min at 4ºC. Following incubation, the plate was washed three times and bound HCR detected using Ultra TMB (Thermo Fisher Scientific) as the HRP substrate. The reaction was terminated by the addition of an equal volume of 0.1 M H2S04. The absorbance at 450 nm was determined using a plate reader (BioTek, Winooski, VT, USA). The apparent equilibrium dissociation constant (Kd) was estimated by fitting the data to a one-site binding model where Y = Bmax × X / (Kd + X) using GraphPad Prism version 6. The goodness of fit was estimated by calculating R2, with values < 0.9 indicating the model was not appropriate. Kd values for individual experiments are reported as the arithmetic mean and standard error of at least three independent trials performed in triplicate (summarized in Table 1).


7

Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 43

Culture of primary rat cortical neurons. Rat embryonic day 18 (E18) cortices (Brainbits, llc, Springfield, Il, USA) were mechanically dissociated and plated in Neurobasal medium supplemented with B-27 (Thermo Fisher Scientific) on poly-D-lysine coated glass cover slips. Half of the culture medium was changed with fresh Neurobasal medium every 4 days starting on day 5. Binding and entry of HCR domains into cortical neurons – indirect immunofluorescence. Neurons (10-14 days in vitro) were washed twice with pre-warmed (37°C) Hank’s Balanced Salt Solution (HBSS) and incubated with indicated concentrations of HCRs in buffer for 5 min at 37°C. Buffers were either high potassium (high K+) to stimulate membrane depolarization (15 mM HEPESNaOH, 95 mM NaCl, 56 mM KCl, 2.2 mM CaCl2, 0.5 mM MgCl2, pH 7.4) or low potassium (low K+) to mimic resting conditions (15 mM HEPES-NaOH, 145 mM NaCl, 5.6 mM KCl, 2.2 mM CaCl2, 0.5 mM MgCl2, pH 7.4). Cells washed three times in HBSS, and fixed in phosphate buffered saline (PBS) containing 4% (w/v) paraformaldehyde, 4% (w/v) sucrose for 30 min at room temperature. Cells were subsequently quenched with 0.1 M glycine in PBS, permeabilized with 0.25% v/v Triton X-100 in PBS, and blocked using Image-IT reagent (Thermo Fisher Scientific). Bound HCR was detected by immunofluorescence using mouse anti-FLAG (clone M2 1:1000, Sigma-Aldrich) and goat anti-mouse IgG Alexa 488 (1:200, Thermo Fisher Scientific). Endogenous synaptophysin was detected using rabbit anti-synaptophysin (clone YE269 1:100, Abcam, Eugene, OR, USA) and anti-rabbit IgG Alexa 568 (1:200, Thermo Fisher Scientific). Cells were processed using standard immunofluorescence procedures, mounted in ProLong Gold Antifade reagent (Thermo Fisher Scientific), and images acquired using a Leica SPE-2 microscope in confocal scanning mode. Fluorescence intensity within individual fields was measured using Leica Application Suite X (Las X). Binding and entry of HCR domains into cortical neurons – Western blotting. Neurons (10-14 days in vitro) were washed twice with pre-warmed HBSS and incubated with indicated concentrations of HCRs in Neurobasal medium for 30 min at 37°C. Next, cells were washed twice with HBSS and further ACS Paragon Plus Environment

8

Page 9 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

incubated for 90 min at 37°C in Neurobasal medium alone. Finally cells were washed twice with icecold HBSS, lysed with Radioimmunoprecipitation assay buffer (RIPA buffer) on ice for 30 min, and lysates clarified by centrifugation at 15000 ×g for 30 min at 4ºC. Cell lysates were combined with ¼ volume of 4× SDS-PAGE sample buffer, resolved on 13.5% (w/v) SDS-Polyacrylamide gels, transferred to PVDF membranes, and subjected to Western blotting. Blots were sequentially probed with mouse anti-FLAG (clone M2 1:10000, Sigma-Aldrich) and goat anti-mouse IgG HRP (1:80000, Thermo Fisher Scientific) and HCRs detected by enhanced chemiluminescence (SuperSignal West Dura, Thermo Fisher Scientific) using a CCD camera system (HD2 Imager, Protein Simple, San Jose, CA, USA). Blots were stripped with Restore PLUS Western Blot Stripping Buffer (Thermo Fisher Scientific) and further probed with a mouse anti-beta actin antibody conjugated to HRP (clone C4 1:5000, Santa Cruz Biotechnology, Inc. Dallas, TX, USA) and detected as above. Cellular intoxication assay.

Botulinum neurotoxins types A and B were purchased from List

Biological Laboratories (Campbell, CA, USA). Neurons (10-14 days in vitro) were washed twice with pre-warmed HBSS and then incubated in high K+ buffer for 5 min at 37°C, in the presence of 5 nM BoNTs with or without the indicated concentrations of HCR competitor proteins. Cells were then washed thrice with HBSS and incubated for an additional 24 h at 37°C in Neurobasal medium. Following treatment, cells were processed for Western blotting as described above. Membranes were probed for SNAP-25 (clone D9A12 1:1000, Cell Signaling Technology, Danvers, MA, USA) or VAMP2 (clone 69.1 1:5000, Synaptic Systems, Goettingen, Germany) as appropriate.

RESULTS HCR/A1 and HCR/A2 bind ganglioside in a similar manner. Evidence suggests that differences in the rates of cellular binding and/or entry may account for the increased toxicity of BoNT/A2 relative to ACS Paragon Plus Environment

9

Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 43

BoNT/A1 in vivo 16-22. One possible mechanism by which the entry of BoNT/A2 could differ is through unique interactions with ganglioside co-receptors

32

. A homology model of HCR/A2 was generated

using the freely available I-TASSER program (http://zhanglab.ccmb.med.umich.edu/services/)

33

with

the co-crystal structure of HCR/A1 bound to GT1b [PDB: 2VU9; 27] as the model template (Figure 1A). Given the high degree of sequence identity shared between the two toxins (~89% for the HCR domain), the overall shape of the HCR/A2 ganglioside binding pocket is preserved, with conservation among residues forming the GBM and those predicted to make hydrogen bonds to either the Sia5 (Ser1275) or Sia6 (Arg1276) sugars. However, HCR/A1 Tyr1117—which is predicted to stabilize binding to Sia5 through hydrophobic packing and the formation of two hydrogen bonds—is replaced by phenylalanine (Phe1117) in HCR/A2 (Figure 1A, green color). Consequently, it was hypothesized that HCR/A2 would display moderately reduced affinity for ganglioside as compared to HCR/A1. In addition, several additional amino acid substitutions in close proximity to the ganglioside binding site were observed, though none of these residues appear likely to interact directly with the carbohydrate moiety (Figure 1A, cyan color).


10

Page 11 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Figure 1. HCR/A1 and HCR/A2 bind gangliosides in a similar manner. (A) Ribbon representation of the HCR/A1 (grey Cα) in complex with a GT1b analog (stick representation, yellow carbon) illustrating the positioning of the oligosaccharide within the ganglioside binding cleft and the intermolecular hydrogen bond interactions (dashed lines). The four bridging water molecules are shown as spheres in red color. All residues making contact with the oligosaccharide are conserved between HCR/A1 and HCR/A2 with the exception of Tyr1117 which is replaced by phenylalanine in HCR/A2 (green). ACS Paragon Plus Environment

11

Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 43

Additional residues in close proximity to the GT1b oligosaccharide that differ between HCR/A1 and HCR/A2 are shown in cyan. (B) Various concentrations of HCR/A1 (●) or HCR/A2 (■) were examined for the ability to bind ganglioside GT1b. (C) Binding of wild-type HCR/A1 (●) and mutated HCR/A1 (A1Y1117A ■; A1Y1117F

▲;

A1S1275A) to ganglioside GD1a. (D) Binding of wild-type HCR/A2 (●) and

mutated HCR/A2 (A2F1117A ■; A1F1117Y

▲;

A2S1275A) to ganglioside GD1a. The ganglioside binding

specificity of wild-type (black symbols) and mutated (white symbols) HCR/A1 (E) or HCR/A2 (F) proteins was compared using gangliosides GD1a (■/□), GD1b (▲/△) and GM1a (▼/▽). All values represent the arithmetic mean and standard error of a minimum of three independent determinations performed in triplicate. Kd values for all HCR derivatives are summarized in Table 1.

Prior to investigating the ganglioside binding properties of BoNT/A2, we confirmed that a recombinant HCR/A2 bound and entered cortical neurons similar to the parental toxin. In agreement with previous observations

12, 32

, binding of HCR/A2 to cortical neurons was enhanced under

depolarizing conditions (high K+ buffer) and co-localized with the synaptic vesicle resident protein synaptophysin (Figure S4), thus supporting the suitability of recombinant HCR/A2 as a surrogate for the native holotoxin. Analysis of the interaction of HCR/A1 and HCR/A2 with a panel of gangliosides demonstrated that both toxins bound to gangliosides GT1b and GD1a with similar affinities (Figure 1B and Table 1), while binding to gangliosides lacking a terminal Sia5 sugar (GD1b and GM1a) was greatly reduced (Figure 1D and 1F; Table 1). Consistent with these data, replacement of HCR/A1 Tyr1117 with the corresponding phenylalanine residue of HCR/A2 or vice-versa did not significantly alter the affinity for GT1b (Figure 1C and 1E). Furthermore, alanine replacement of Ser1275 in either HCR/A1 or /A2 did not noticeably affect ganglioside binding, suggesting that hydrogen bonds formed by either Tyr1117 or


12

Page 13 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Ser1275 do not play critical roles in the recognition of Sia5 (Figure 1C and 1E; Table 1). Indeed, alanine replacement of the tyrosyl or phenyl rings of HCR/A1 and HCR/A2 respectively, increased binding affinity to gangliosides GT1b and GD1a by an order of magnitude, without altering ganglioside binding specificity ((Figure 1D and 1F; Table 1). The observation that alanine replacement of Tyr1117 increases affinity towards GT1b and GD1a has been previously reported

34

, and can now also be

extended to alanine replacement of Phe1117 in HCR/A2. HCR/A1Y1117A and HCR/A2F1117A variants enter cells more rapidly than wild-type proteins. The prior invention disclosure also demonstrated that replacement of Tyr1117 with Ala, Cys, or Val all resulted in a 3-fold increase in overall neurotoxicity

34

. Therefore, the functional consequences of

Phe1117 replacement in A2 were investigated. Consistent with previous reports, the entry of HCR/A2 relative to HCR/A1 was elevated at low concentrations of toxin (Figure 2A) 32, though total uptake of both proteins was similar at ~ 10 µM concentration (data not shown). In agreement with the in vitro binding data, both HCR/A1Y1117A and HCR/A2F1117A bound and entered neurons ~2-4-fold more efficiently than the parental HCR proteins as determined by immunofluorescence assay (Figure 2A and 2B), and was further substantiated by Western blotting of cell lysates from intoxicated neurons (Figure 3A). Next, the ability of HCR derivatives to inhibit the intracellular activity of native BoNT/A1 holotoxin in cortical neurons was tested. As reported previously, cleavage of SNAP25 by BoNT/A1 was inhibited in a dose dependent fashion by HCR/A1, and to a greater efficiency by co-incubation with either HCR/A1Y1117A or HCR/A2 (Figure 3B)

32, 35, 36

. HCR/A2F1117A was the most effective inhibitor,

being ~10-fold more efficient than HCR/A1 (Figure 3B). Recently, Barbieri and co-workers proposed that a higher degree of receptor occupancy, rather than rates of internalization, may account for the increased cellular toxicity of BoNT/A2 relative to BoNT/A1

32

. It was hypothesized that regions

surrounding the ganglioside binding pocket may contribute to affinity, without altering the specificity of HCR/A for gangliosides. Yet, a HCR/A1 variant (QFN1254-1256LYD; IERS1271-1274VGKA)


13

Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 43

containing the corresponding residues from HCR/A2 bound GT1b in vitro with similar affinity to the wild-type HCR/A1 (Figure S5). While this suggests that non-conserved residues surrounding the ganglioside binding pocket do not directly influence ganglioside binding, we cannot exclude the possibility that such changes between A1 and A2 alter the affinity of the toxins for the membrane bilayer.

Figure 2. HCR/A1Y1117A and HCR/A2F1117A variants enter cells more rapidly than wild type proteins. (A) Primary rat cortical neurons were incubated with the indicated HCR domains [50 nM] for 5 min at 37°C in high K+ buffer to stimulate vesicle recycling. Cells were washed,

fixed,

permeabilized,

and

subjected

to

immunostaining analysis. Binding of HCR derivatives was

detected

using

an

anti-FLAG

antibody.

Synaptophysin was labeled as a marker for presynaptic terminals. The scale bar represents 10 µm in all panels. (B) Quantification of relative binding by individual HCR domains. All values represent the arithmetic mean and standard error of fluorescence intensity from 15 random fields and is representative of 3 biological repeats.


14

Page 15 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Figure 3. HCR derivatives inhibit the activity of BoNT/A1 in primary cortical neurons. (A) Left panel, primary rat cortical neurons were exposed to HCRs (indicated concentrations) for 30 min in Neurobasal medium. Cells were washed and further incubated in HCR-free media for 90 min to allow for HCR internalization. Cell lysates (25 µg total protein) were subjected to immunoblot analysis. Cells that were not exposed to toxins served as the control (No toxin). Beta-actin was detected as an internal ACS Paragon Plus Environment

15

Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 43

control for loading of cell lysates. Right panel, HCR binding was quantified by densitometry using AlphaView software and normalized to total beta-actin expression. Densitometry values (arbitrary units) represent the arithmetic mean and standard error of 5 independent experiments. (B) Left panel, cells were exposed to BoNT/A (5 nM, 5 min) under depolarizing conditions (High K+ buffer). Cells were then washed and incubated for an additional 24 h in medium alone. Cell lysates (30 µg total protein) were subjected to immunoblot analysis. Cells that were not exposed to BoNT/A1 served as the control (No toxin), while beta-actin serves as a loading control. Cleavage of SNAP-25 by BoNT/A yielded a smaller fragment that is indicated by an asterisk. Right panel, Cleavage of SNAP-25 was quantified by densitometry using AlphaView software. Values represent the arithmetic mean and standard error of 5 independent experiments.

A conserved hydrophobic residue plays a key role in recognition of Sia5. Previous independent studies have argued that BoNT types A, B, E, and F preferentially bind to gangliosides containing Sia5 while TeNT is unique in binding to gangliosides in a Sia5-independent manner

25, 37, 38

. In agreement

with these observations, comparative analysis revealed that HCR domains of BoNT serotypes A, B, E, and F and TeNT all bound to ganglioside GD1a (Sia5), albeit with moderate differences in binding affinity, while only TeNT efficiently bound to ganglioside GM1a (no Sia5, Figures 4A and 4B). This leads to the question of why BoNTs and TeNT differ in their requirement for the Sia5 sugar. Superposition of the HCR/A1-GT1b, HCR/A1-GD1a, HCR/B-GD1a, HCR/F-GD1a, and HCR/TGT1b analog complexes show that the overall interaction with ganglioside is similar among the toxins (PDB numbers 2VU9, 5TPC, 4KBB, 3RSJ, and 1FV2 respectively). In each case, the GalNAc3-Gal4 moiety is bound at the base of the binding pocket, forming extensive interactions with residues forming the GBM. In contrast, there are large differences at the top of the ganglioside binding cleft, wherein


16

Page 17 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

BoNTs contain a series of semi-conserved residues that form a more closed Sia5 binding site. In particular, the BoNTs all contain a hydrophobic residue (phenylalanine in BoNTs A, E, and F; isoleucine in BoNTs B, C, and D) located at the interface of the GBM and the Sia5 pocket that is replaced by threonine in TeNT (Figure S6).

Figure 4. Wild type HCR binding kinetics. Various concentrations of HCR/A1 (●), HCR/B (■), HCR/E (▲), HCR/F (▼) and HCR/T (◆) were examined for their ability to bind ganglioside GD1a (A) or GM1a (B). All values represent the arithmetic mean and standard deviation of at least four independent experiments performed in triplicate. Kd values for all HCR derivatives are summarized in Table 1.

Substitution of the hydrophobic residue of HCR/A (Phe1252) or HCR/F (Phe1240) with the corresponding isoleucine residue of HCR/B reduced ganglioside binding by 1-2 orders of magnitude, while replacement with either alanine (Figure 5A and 5B) or threonine (Table 1) abolished ganglioside


17

Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 43

binding. It was not possible to test the role of Phe1214 in HCR/E as mutation of this residue to either alanine, threonine or isoleucine caused the protein to aggregate in solution (data not shown). Next, the effect of mutating HCR/F Phe1240 to alanine was tested in a previously described HCR/F variant (H1241K) capable of binding to gangliosides in a Sia5 independent manner

25

. Consistent with these

observations, HCR/FH1241K bound with high affinity to GD1a and, to a lesser extent, GM1a. In contrast, while HCR/FH1241K, F1240A bound GM1a with similar affinity to HCR/FH1241K, binding to GD1a was not observed (Figure 5C). Replacement of isoleucine1239 with phenylalanine increases binding and entry of HCR/B. While substitution of Ile1239 with alanine or threonine abolished binding to ganglioside GD1a, introduction of the corresponding phenylalanine residue of HCR/A1 and HCR/F increased binding by an order of magnitude (Figure 6A and Table 1). Comparative analysis of HCR/B and HCR/BI1239F binding to additional gangliosides demonstrated both wild-type and mutated HCR/B displayed similar specificity, preferentially recognizing gangliosides containing a terminal Gal4-Sia5 linkage (Figure 6A and 6B). As an aside, it is noted that the relative affinities of HCR/A1 and HCR/B for GT1b reported herein are similar (483 nM vs. 470 nM, respectively). This observation differs from the previous work of Binz and colleagues who reported that HCR/B bound GT1b with significantly higher affinity than HCR/A1

23

.

The source of this discrepancy cannot be explained at this time.


18

Page 19 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Figure 5. A conserved hydrophobic residue facilitates binding at the Sia5 position. Various concentrations of HCR domains were analyzed for their ability to bind GD1a (dark symbols) or GM1a (white symbols) as indicated: (A) HCR/A1 (●) or HCR/A1 Phe1252 variants (HCR/A1F1252I ■, HCR/A1F1252A ▲); (B) HCR/F (●) or HCR/F Phe1240 variants (HCR/FF1240I ■, HCR/FF1240A ▲); and (C) HCR/FH1241K (●/○) or HCR/FH1241K, F1240A (■/□) variant. All values represent the arithmetic mean and standard error of a minimum of three independent determinations performed in triplicate. Kd values for all HCR derivatives are summarized in Table 1. ACS Paragon Plus Environment

19

Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 43

Figure 6. HCR/BI1239F binds gangliosides with greater affinity than wild-type HCR/B. (A) HCR/B (●) or HCR/B Ile1239 variants (HCR/BI1239F ■, HCR/BI1239A ▲) were analyzed for their ability to bind GD1a. (B) The ganglioside binding specificity of HCR/B (black symbols) or HCR/BI1239F (white symbols) was compared using gangliosides GT1b (●/○), GD1b (■/□) and GM1a (▲/△). All values represent the arithmetic mean and standard error of a minimum of three independent determinations performed in triplicate. Kd values for all HCR derivatives are summarized in Table 1.

Next, the biological activity of HCR/BI1239F was investigated by monitoring both uptake into neurons and the ability of HCR derivatives to compete with native BoNT/B holotoxin. The binding and entry of HCR/B and HCR/BI1239F into primary cortical neurons was enhanced under depolarizing conditions (High K+ buffer, Figure 7A) and is consistent with prior studies which demonstrated that entry of BoNT/B is mediated by synaptotagmins I and II

39, 40

. In accord with the ganglioside binding data,

binding and entry of HCR/BI1239F was enhanced ~10-fold relative to wild-type HCR/B as judged by ACS Paragon Plus Environment

20

Page 21 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

immunofluorescence assay and Western blotting (Figure 7A and 7B). A competition assay showed that HCR/BI1239F competed with native BoNT/B for neuronal binding more effectively that wild-type HCR/B (Figure 7C), signifying that introduction of the Ile1239Phe mutation into BoNT/B would increase toxin potency. Collectively, the data presented in Figures 5-7 suggest the semi-conserved hydrophobic

residue within the BoNTs plays an important role in the recognition of Sia5. Figure 7. HCR/BI1239F binds and enters cells more efficiently than wild-type HCR/B. (A) Left panel, primary rat cortical neurons were exposed to HCR/B or HCR/BI1239F (50 nM, 5 min) in either low K+ ACS Paragon Plus Environment

21

Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 43

buffer (to mimic the resting state) or high K+ buffer to stimulate vesicle recycling. Cells were washed, fixed, permeabilized, and subjected to immunostaining analysis. Binding of HCR derivatives was detected using an anti-FLAG antibody. Synaptophysin was labeled as a marker for presynaptic terminals. The scale bar represents 10 µm in all panels. Right panel, quantification of relative binding by HCR domains at the indicated concentrations. HCR concentration in low K+ buffer was 50 nM. All values represent the arithmetic mean and standard error of fluorescence intensity from 15 random fields and is representative of 2 biological repeats. (B) Left panel, primary rat cortical neurons were exposed to HCRs (indicated concentrations) for 30 min in Neurobasal medium. Cells were washed and further incubated in HCR-free media for 90 min to allow for HCR internalization. Cell lysates (25 µg total protein) were subjected to immunoblot analysis. Cells that were not exposed to HCR served as the control (No toxin). Beta-actin was detected as an internal control for loading of cell lysates. Right panel, HCR binding was quantified by densitometry using AlphaView software and normalized to total betaactin expression. Densitometry values (arbitrary units) represent the arithmetic mean and standard error of 5 independent experiments. (C) Left panel, cells were exposed to BoNT/B (5 nM, 5 min) under depolarizing conditions (High K+ buffer). Cells were then washed and incubated for an additional 24 h in medium alone. Cell lysates (30 µg total protein) were subjected to immunoblot analysis. Cells that were not exposed to BoNT/B served as the control (No toxin), while beta-actin serves as a loading control. Cleavage of VAMP2 by BoNT/B results in a loss of detectable signal on Western blotting. Right panel, Cleavage of VAMP2 was quantified by densitometry using AlphaView software and expressed relative to no toxin control. Values represent the arithmetic mean and standard error of 5 independent experiments.

Engineering of a TeNT variant dependent on Sia5 for ganglioside binding. The observation that TeNT binds to gangliosides GD1a and GM1a with near identical affinity argues that Sia5 does not ACS Paragon Plus Environment

22

Page 23 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

contribute to interaction in a significant way. Therefore, we reasoned that by altering the Sia5 binding pocket of TeNT to more closely resemble that of the BoNTs, it should be possible to induce interactions with Sia5. However, initial replacement of Thr1270 with phenylalanine completely ablated binding to both GD1a and GM1a (Figure 8A). As a result, additional substitutions were introduced into TeNT designed to increase interactions with Sia5. Firstly, Ser1135 was replaced with arginine, which in the HCR/F-GD1a complex forms a hydrogen bond to Sia5 in addition to forming hydrophobic interactions between the aliphatic region of the side chain and the sugar ring. At the same time, Ala1134 was replaced with the corresponding leucine from HCR/F to help support the positioning of the Arg1135 side chain as well as to increase the overall hydrophobicity of the region. As shown in Figure 8B, the introduced substitutions (HCR/TA1134L,

S1135R, T1270F

) restored binding to GM1a and caused a modest

increase in affinity for GD1a as compared to wild type TeNT. Figure 8. A TeNT variant that binds specifically to gangliosides containing a Sia5 moiety. Various concentrations of HCR derivatives were analyzed for their ability to bind gangliosides GD1a (black ACS Paragon Plus Environment

23

Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 43

symbols) or GM1a (white symbols) as indicated: (A) HCR/T (●) or HCR/TT1270F (■/□); (B) HCR/TA1134L,

S1135R, T1270F

(▲/△); (C) HCR/T (●) or HCR/T-A

Loop

(▼); and (D) HCR/T (●/○) or

HCR/TA1134L, S1135R, T1270F-A Loop (◆/◇). All values represent the arithmetic mean and standard error of a minimum of three independent determinations performed in triplicate. Kd values for all HCR derivatives are summarized in Table 1.

BoNT/F makes an additional interaction with Sia5 through a non-conserved arginine residue (Arg1256). While direct replacement of the corresponding position was considered, differences in the orientation of the loop linking α-helix 4 with β-strand 30 (residues 1294-1302) of HCR/T relative to HCR/F suggested this was not feasible. Therefore, it was decided to replace the entire loop region of HCR/T with the equivalent sequence of HCR/F. However, HCR/TA1134L,

S1135R, T1270F

-FLoop and

HCR/TA1134L, S1135R, T1270F did not differ in affinity for gangliosides GD1a or GM1a (data not shown). Based on this observation, additional exchanges were performed using the corresponding loop regions of BoNTs A (residues 1270-1279), B (residues 1267-1279), and E (residues 1228-1237). Among these chimeric proteins, HCR/TA1134L,

S1135R, T1270F

-ALoop variant was stable upon purification and therefore

suitable for further investigation. Unexpectedly, while HCR/TA1134L, S1135R, T1270F-ALoop bound to GD1a with similar affinity as HCR/T, binding to GM1a was not observed (Figure 8C). The reason for the specific loss of GM1a binding is not clear. However, insertion of the A loop alone is sufficient to disrupt binding to ganglioside (Figure 8D), suggesting that the overall structure of the Sia5 binding site plays a key role in regulating binding to gangliosides.

DISCUSSION


24

Page 25 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Clostridial neurotoxins (CNTs) bind and enter nerve terminals via interactions with synaptic vesicle resident proteins and an array of complex gangliosides present on the presynaptic membrane 10, 36, 39, 4150

. In this classical double receptor model 51, the carbohydrate components of gangliosides are suggested

to act as the initial attachment sites for the toxin due to their high local concentration on the presynaptic membrane and high lateral mobility. The carbohydrate moiety itself binds in a shallow groove formed by the ganglioside binding motif (GBM; E/D…H/K…SXWY…G), located at the distal tip of the HCR domain

25-28

. While this motif is essential for ganglioside recognition, it does not fully account for

differences in both the affinity and specificity of individual CNTs towards a subset of gangliosides 25, 27, 28

. In this study, we aimed to understand the basis by which CNTs discriminate between the

carbohydrate moieties of individual gangliosides. Our data support a model in which recognition of Sia5 by BoNTs is driven primarily through hydrophobic interactions between the protein and the sugar moiety. A recent study by Stenmark, Widmalm, and co-workers identified the terminal carbohydrate branch of GD1a ([α-Neu5Ac-(2→3)]-β-Gal-(1→3)-β-GalNAc-) as the active moiety that binds with a high degree of propensity to BoNT/A1

52

. Molecular docking simulations using carbohydrate fragments of GD1a

revealed that a galacto-configured sugar (Gal or GalNAc) preferentially binds at a subsite formed by the conserved His1253 and Trp1266 residues of the GBM (denoted as subsite B, Figure S7A). Ligands carrying a terminal Sia-α(2→3)-Gal moiety also occupy site B, with the Sia sugar largely occupying the adjacent A subsite largely defined by Tyr1117 and Phe1252 (denoted site A in Figure S7A)

52

. These

findings are consistent with earlier studies of ganglioside binding by BoNT/A1 38, 53, 54, in addition to the data provided here for HCR/A1 and HCR/A2 (summarized in Table 1). Similarly, HCR/B, HCR/E, and HCR/F, but not HCR/T, also show specificity towards gangliosides containing a terminal Sia5-Gal4 moiety (i.e., GT1b and GD1a, Figures 4 and 6, Table 1). Thus these data further support the model of ACS Paragon Plus Environment

25

Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 43

Stenmark, Widmalm, and co-workers, wherein the carbohydrate moiety of gangliosides is responsible for determining specificity of binding. Given the high degree of sequence and structural conservation in the ganglioside binding pockets of TeNT and the BoNTs, how does one account for the observed differences in specificity of binding? Insight into this question has emerged in recent years with the determination of the solution structures of several HCR domains in complex with analogs of ganglioside carbohydrate moieties

25-28

. It was

previously reported that substitution of two arginine residues in HCR/F (Arg1111 and Arg1256)—which form hydrogen bonds to Sia5—reduced affinity towards GD1a by ~70-fold

25

. While these arginine

residues are not conserved, structural data suggested other BoNT serotypes possessed additional residues located outside of the GBM that performed analogous functions. In particular, structural data suggested HCR/A1 Tyr1117 and Ser1275 form hydrogen bonds (two and one, respectively) to Sia5 which, in conjunction with additional hydrophobic interactions, were proposed to stabilize binding to GT1b

27

. However, mutational analysis of HCR/A1 and HCR/A2 suggested that both Tyr1117 and

Ser1275 make only minimal contributions to ganglioside binding (Figure 1 and Table 1). This is further supported by the recent crystal complex data showing different hydrogen bonding networks for Sia5 between the HCR/A●GT1b, HCR/A1●GD1a, and HCR/A1●sialyl-T antigen complexes 27, 52. Though it cannot be excluded that packing of Sia5 against the tyrosyl ring of HCR/A1 Y1117 is important for binding, this argument is weakened by the observation that alanine replacement improved affinity for GT1b and GD1a in vitro, accelerated uptake into neurons, and increased neuronal toxicity, as judged by competition assay (Figures 1, 2, and 3). These data are in agreement with a previously described BoNT/A1Y1117A variant and can now be extended to BoNT/A2

34

. While the mechanistic basis for this

change is unclear, it should be noted that in both the HCR/A●GT1b and HCR/A1●GD1a complexes Tyr1117 rotates around the C-beta C-gamma bond to accommodate Sia5 upon binding

27, 52

; this may

suggest that the interaction of the tyrosyl side chain with the Sia5 group is energetically unfavorable and


26

Page 27 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

therefore reduces overall binding affinity. Indeed, modeling of the HCR/A1Y1117A variant suggests the removal of the ring structure expands the ganglioside binding pocket, potentially providing lower constraints on the conformations Sia5 can adopt within the A subsite (Figure S7B). The observation that alanine replacement of HCR/F Arg1111 and Arg1256 reduced affinity towards GD1a without ablating binding 25, suggested that hydrophobic interactions between the protein and Sia5 could be important to stabilize ganglioside binding. Structural alignment of CNT HCR domains identified a semi-conserved hydrophobic residue within the BoNTs (Phe in BoNTs A, E, and F; Ile in BoNTs B, C, and D) that shapes the A subsite of the ganglioside binding pocket. Replacement of the phenylalanine residues of HCR/A1 and HCR/F (Phe1252 and Phe1240 respectively) with the corresponding isoleucine residue of HCR/B caused a 1-2 log decrease in binding to GD1a, while substitution with alanine abolished binding (Figures 5A and 5B). While alanine replacement of Ile1239 in HCR/B caused a similar ablation of binding, HCR/BI1239F bound GT1b and GD1a ~10-fold more efficiently than wild-type HCR/B (Figure 6 and Table 1). Similar to HCR/A1Y1117A, the increased affinity of HCR/BI1239F for gangliosides in vitro also resulted in accelerated entry into neurons and greater toxicity as judged by the ability to compete with BoNT/B holotoxin for binding and entry (Figure 7). CD spectroscopy of wild-type and mutated HCR variants revealed only minor differences in the respective spectra, arguing that changes in ganglioside binding were not the result of major changes in protein secondary structure (Figure S3). These data argue that BoNT/BI1239F will possess greater toxicity in vivo, potentially expanding its clinical efficacy in the treatment of neurologic conditions. Furthermore, these observations highlight the need to fully characterize the biological activities of subtype variants, as individual amino acid changes can have unexpected effects on toxin activity. Finally, a variant of TeNT was engineered that was dependent on Sia5 for ganglioside binding. While HCR/TT1270F displayed a large decrease in affinity for ganglioside, further substitutions designed to increase the overall hydrophobicity of the A subsite (Ala1134Leu and Ser1135Arg) restored binding to ACS Paragon Plus Environment

27

Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 43

near wild-type levels (Figures 8A and 8B). Thus, while the conserved hydrophobic residue present in the BoNTs appears to play a key role in the recognition of Sia5, it does not appear to be the sole determinant of ganglioside binding specificity. In agreement with this statement, was the observation that replacement of the short loop linking α-helix 4 with β-helix 30 (residues 1294-1302) of HCR/T with the corresponding region from HCR/A1 was sufficient to disrupt the ability to bind ganglioside GM1a without affecting binding to GD1a (Figure 8D). In summary, these studies provide new insights into the mechanism of ganglioside recognition by individual clostridial neurotoxins and suggest that the specificity of ganglioside binding is determined in large part through the interaction of the Sia5 sugar with the hydrophobic A subsite of the ganglioside binding pocket. Understanding the mechanisms of ganglioside recognition opens up the possibility of generating novel toxin variants with improved and/or expanded clinical potential.


28

Page 29 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Table 1. Relative affinities of HCR derivatives for the indicated gangliosides.

HCR/A1 HCR/A1Y1117A HCR/A1Y1117F HCR/A1S1275A HCR/A1F1252I HCR/A1F1252A HCR/A1F1252T

GT1b 483 ± 29 38 ± 7 263 ± 17 208 ± 16 n.d. n.d. n.d.

HCR/A2 HCR/A2F1117A HCR/A2F1117Y HCR/A2S1275A

355 ± 25 20 ± 2 258 ± 28 335 ± 29

(9) (6) (6) (6)

565 ± 47 68 ± 5 n.d. n.d.

(9) (6)

>10,000 >10,000 n.d. n.d.

(3) (3)

>10,000 >10,000 n.d. n.d.

(3) (3)

HCR/B HCR/BI1239F HCR/BI1239A HCR/BI1239T

470 ± 28 34 ± 3 n.d. n.d.

(9) (3)

1111 ± 85 75 ± 7 >10,000 >10,000

(12) (6) (3) (3)

>10,000 (3) 9833 ± 546 (3) n.d. n.d.

>10,000 >10,000 >10,000 >10,000

(3) (3) (3) (3)

HCR/E

242 ± 23

(2)

219 ± 12

(6)

>10,000

(2)

>10,000

(3)

HCR/F HCR/FF1240I HCR/FF1240A HCR/FF1240T HCR/FH1241K HCR/FH1241K, F1240A

57 ± 3 n.d. n.d. n.d. n.d. n.d.

(2)

63 ± 3 349 ± 40 >10,000 >10,000 6 ± 0.3 >10,000

(9) (3) (3) (3) (6) (3)

>10,000 n.d. n.d. n.d. n.d. n.d.

(2)

>10,000 >10,000 >10,000 >10,000 579 ± 52 2042 ± 177

(3) (3) (3) (3) (6) (6)

HCR/T† HCR/TT1270F HCR/TT1270A HCR/T A1134L, S1135R,

234 ± 48 n.d. n.d. n.d.

(3)

57 ± 3 6775 ± 771 61 ± 5 13 ± 1

(22) (3) (2) (6)

322 ± 21 n.d. n.d. n.d.

(3)

24 ± 2 4083 ± 579 28 ± 2 78 ± 7

(22) (3) (2) (6)

(12) (6) (6) (6)

GD1a 522 ± 26 48 ± 5 n.d. n.d. >10,000 >10,000 >10,000

Apparent Kd (nM)* GD1b (18) 8232 ± 487 (9) (6) >10,000 (6) n.d. n.d. (3) n.d. (3) n.d. (3) n.d.

GM1a >10,000 >10,000 n.d. n.d. >10,000 >10,000 >10,000

(9) (6)

(3) (3) (3)

T1270F

HCR/T A1134L, S1135R, n.d. 52 ± 5 (4) n.d. >10,000 (4) T1270F -A Loop * all values represent the arithmetic mean and standard error of n (number of experiments shown in brackets) independent experiments performed in triplicate. ACS Paragon Plus Environment

29

Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 43

n.d. not determined. † HCR/T (contains R1226L mutation to inactivate the second (non-conserved) ganglioside binding site.

ASSOCIATED CONTENT Supporting Information. 7 figures. Figures show a comparison of the protein-carbohydrate interactions of botulinum and tetanus neurotoxins, an analysis of HCR purity by SDS polyacrylamide gel electrophoresis, CD spectra of wild-type and mutated HCR derivatives, entry of HCR/A2 into primary rat cortical neurons, binding data for a HCR/A1 variant (QFN1254-1256LYD; IERS1271-1274VGKA), a structure-based sequence alignment of clostridial neurotoxins, and a cartoon illustrating the location of the ganglioside binding site of HCR/A1 .

AUTHOR INFORMATION Corresponding Author * Address correspondence to Michael R. Baldwin, 1 Hospital Drive, M618 Medical Sciences Building, Columbia, MO, 65212, (Telephone- +1-573-884-2915; Fax- +1-573-882-4287), [email protected] Present Address † Department of Microbiology and Immunology, The University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma, USA Author Contributions


30

Page 31 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

The manuscript and experiments performed herein was through contributions of all authors. All authors have given approval to the final version of the manuscript. # These authors contributed equally. Funding Sources This research was supported by funds provided by The University of Missouri-Columbia.

ABBREVIATIONS

TeNT, tetanus neurotoxin; BoNT, botulinum neurotoxin; CNT, clostridial neurotoxin; HC, heavy chain; HCT, heavy chain translocation; HCR, heavy chain receptor; LC, light chain; SNARE, soluble Nethylmaleimide-sensitive fusion protein attachment protein receptor; GBM, ganglioside binding motif; Gal, galactose, GalNAc, N-acetylgalactosamine; Neu5Ac, N-acetylneuraminic acid; PBS, phosphate buffered saline; HBSS, Hank’s balanced salt solution.

REFERENCES

[1] Schiavo, G., Matteoli, M., and Montecucco, C. (2000) Neurotoxins affecting neuroexocytosis, Physiological reviews 80, 717-766. [2] Johnson, E. A. (1999) Clostridial toxins as therapeutic agents: benefits of nature's most toxic proteins, Annu Rev Microbiol 53, 551-575. [3] Williamson, C. H., Sahl, J. W., Smith, T. J., Xie, G., Foley, B. T., Smith, L. A., Fernandez, R. A., Lindstrom, M., Korkeala, H., Keim, P., Foster, J., and Hill, K. (2016) Comparative genomic analyses reveal broad diversity in botulinum-toxin-producing Clostridia, BMC Genomics 17, 180. [4] Kull, S., Schulz, K. M., Weisemann, J., Kirchner, S., Schreiber, T., Bollenbach, A., Dabrowski, P. W., Nitsche, A., Kalb, S. R., Dorner, M. B., Barr, J. R., Rummel, A., and Dorner, B. G. (2015) Isolation and functional characterization of the novel Clostridium botulinum neurotoxin A8 subtype, PloS one 10, e0116381. [5] Montecucco, C., and Rasotto, M. B. (2015) On botulinum neurotoxin variability, mBio 6. [6] Hill, K. K., Xie, G., Foley, B. T., and Smith, T. J. (2015) Genetic diversity within the botulinum neurotoxin-producing bacteria and their neurotoxins, Toxicon : official journal of the International Society on Toxinology 107, 2-8.


31

Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 43

[7] Smith, T. J., Hill, K. K., and Raphael, B. H. (2015) Historical and current perspectives on Clostridium botulinum diversity, Res Microbiol 166, 290-302. [8] Lacy, D. B., Tepp, W., Cohen, A. C., DasGupta, B. R., and Stevens, R. C. (1998) Crystal structure of botulinum neurotoxin type A and implications for toxicity, Nat Struct Biol 5, 898-902. [9] Montal, M. (2010) Botulinum neurotoxin: a marvel of protein design, Annu Rev Biochem 79, 591617. [10] Binz, T., and Rummel, A. (2009) Cell entry strategy of clostridial neurotoxins, Journal of neurochemistry 109, 1584-1595. [11] Binz, T. (2013) Clostridial neurotoxin light chains: devices for SNARE cleavage mediated blockade of neurotransmission, Current topics in microbiology and immunology 364, 139-157. [12] Pier, C. L., Chen, C., Tepp, W. H., Lin, G., Janda, K. D., Barbieri, J. T., Pellett, S., and Johnson, E. A. (2011) Botulinum neurotoxin subtype A2 enters neuronal cells faster than subtype A1, FEBS letters 585, 199-206. [13] Wang, D., Krilich, J., Pellett, S., Baudys, J., Tepp, W. H., Barr, J. R., Johnson, E. A., and Kalb, S. R. (2013) Comparison of the catalytic properties of the botulinum neurotoxin subtypes A1 and A5, Biochimica et biophysica acta 1834, 2722-2728. [14] Whitemarsh, R. C., Tepp, W. H., Bradshaw, M., Lin, G., Pier, C. L., Scherf, J. M., Johnson, E. A., and Pellett, S. (2013) Characterization of botulinum neurotoxin A subtypes 1 through 5 by investigation of activities in mice, in neuronal cell cultures, and in vitro, Infection and immunity 81, 3894-3902. [15] Pellett, S., Tepp, W. H., Whitemarsh, R. C., Bradshaw, M., and Johnson, E. A. (2015) In vivo onset and duration of action varies for botulinum neurotoxin A subtypes 1-5, Toxicon : official journal of the International Society on Toxinology 107, 37-42. [16] Torii, Y., Goto, Y., Nakahira, S., Kozaki, S., Kaji, R., and Ginnaga, A. (2015) Comparison of Systemic Toxicity between Botulinum Toxin Subtypes A1 and A2 in Mice and Rats, Basic Clin Pharmacol Toxicol 116, 524-528. [17] Koizumi, H., Goto, S., Okita, S., Morigaki, R., Akaike, N., Torii, Y., Harakawa, T., Ginnaga, A., and Kaji, R. (2014) Spinal Central Effects of Peripherally Applied Botulinum Neurotoxin A in Comparison between Its Subtypes A1 and A2, Front Neurol 5, 98. [18] Mukai, Y., Shimatani, Y., Sako, W., Asanuma, K., Nodera, H., Sakamoto, T., Izumi, Y., Kohda, T., Kozaki, S., and Kaji, R. (2014) Comparison between botulinum neurotoxin type A2 and type A1 by electrophysiological study in healthy individuals, Toxicon : official journal of the International Society on Toxinology 81, 32-36. [19] Akaike, N., Shin, M. C., Wakita, M., Torii, Y., Harakawa, T., Ginnaga, A., Kato, K., Kaji, R., and Kozaki, S. (2013) Transsynaptic inhibition of spinal transmission by A2 botulinum toxin, J Physiol 591, 1031-1043. [20] Ma, L., Nagai, J., Sekino, Y., Goto, Y., Nakahira, S., and Ueda, H. (2012) Single application of A2 NTX, a botulinum toxin A2 subunit, prevents chronic pain over long periods in both diabetic and spinal cord injury-induced neuropathic pain models, J Pharmacol Sci 119, 282-286. [21] Torii, Y., Akaike, N., Harakawa, T., Kato, K., Sugimoto, N., Goto, Y., Nakahira, S., Kohda, T., Kozaki, S., Kaji, R., and Ginnaga, A. (2011) Type A1 but not type A2 botulinum toxin decreases the grip strength of the contralateral foreleg through axonal transport from the toxin-treated foreleg of rats, J Pharmacol Sci 117, 275-285. [22] Torii, Y., Kiyota, N., Sugimoto, N., Mori, Y., Goto, Y., Harakawa, T., Nakahira, S., Kaji, R., Kozaki, S., and Ginnaga, A. (2011) Comparison of effects of botulinum toxin subtype A1 and A2 using twitch tension assay and rat grip strength test, Toxicon : official journal of the International Society on Toxinology 57, 93-99.


32

Page 33 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

[23] Rummel, A., Mahrhold, S., Bigalke, H., and Binz, T. (2004) The HCC-domain of botulinum neurotoxins A and B exhibits a singular ganglioside binding site displaying serotype specific carbohydrate interaction, Molecular microbiology 51, 631-643. [24] Rummel, A., Bade, S., Alves, J., Bigalke, H., and Binz, T. (2003) Two carbohydrate binding sites in the H(CC)-domain of tetanus neurotoxin are required for toxicity, Journal of molecular biology 326, 835-847. [25] Benson, M. A., Fu, Z., Kim, J. J., and Baldwin, M. R. (2011) Unique ganglioside recognition strategies for clostridial neurotoxins, The Journal of biological chemistry 286, 34015-34022. [26] Fotinou, C., Emsley, P., Black, I., Ando, H., Ishida, H., Kiso, M., Sinha, K. A., Fairweather, N. F., and Isaacs, N. W. (2001) The crystal structure of tetanus toxin Hc fragment complexed with a synthetic GT1b analogue suggests cross-linking between ganglioside receptors and the toxin, The Journal of biological chemistry 276, 32274-32281. [27] Stenmark, P., Dupuy, J., Imamura, A., Kiso, M., and Stevens, R. C. (2008) Crystal structure of botulinum neurotoxin type A in complex with the cell surface co-receptor GT1b-insight into the toxin-neuron interaction, PLoS pathogens 4, e1000129. [28] Berntsson, R. P., Peng, L., Dong, M., and Stenmark, P. (2013) Structure of dual receptor binding to botulinum neurotoxin B, Nature communications 4, 2058. [29] Baldwin, M. R., Tepp, W. H., Pier, C. L., Bradshaw, M., Ho, M., Wilson, B. A., Fritz, R. B., Johnson, E. A., and Barbieri, J. T. (2005) Characterization of the antibody response to the receptor binding domain of botulinum neurotoxin serotypes A and E, Infection and immunity 73, 6998-7005. [30] Chen, C., Fu, Z., Kim, J. J., Barbieri, J. T., and Baldwin, M. R. (2009) Gangliosides as high affinity receptors for tetanus neurotoxin, The Journal of biological chemistry 284, 26569-26577. [31] Chen, C., Baldwin, M. R., and Barbieri, J. T. (2008) Molecular basis for tetanus toxin coreceptor interactions, Biochemistry 47, 7179-7186. [32] Kroken, A. R., Blum, F. C., Zuverink, M., and Barbieri, J. T. (2017) Entry of Botulinum Neurotoxin Subtypes A1 and A2 into Neurons, Infection and immunity 85, e00795-816. DOI: 10.1128/IAI.00795-16 [33] Roy, A., Kucukural, A., and Zhang, Y. (2010) I-TASSER: a unified platform for automated protein structure and function prediction, Nat Protoc 5, 725-738. [34] Rummel, A. (October 23, 2012) Transport Protein Which Is Used To Introduce Chemical Compounds Into Nerve Cells. U.S. Patent In USPTO (USPTO, Ed.), IPSEN BIOINNOVATION Ltd United States of America. [35] Baldwin, M. R., and Barbieri, J. T. (2007) Association of botulinum neurotoxin serotypes a and B with synaptic vesicle protein complexes, Biochemistry 46, 3200-3210. [36] Fu, Z., Chen, C., Barbieri, J. T., Kim, J. J., and Baldwin, M. R. (2009) Glycosylated SV2 and gangliosides as dual receptors for botulinum neurotoxin serotype F, Biochemistry 48, 5631-5641. [37] Kitamura, M., Iwamori, M., and Nagai, Y. (1980) Interaction between Clostridium botulinum neurotoxin and gangliosides, Biochimica et biophysica acta 628, 328-335. [38] Ochanda, J. O., Syuto, B., Ohishi, I., Naiki, M., and Kubo, S. (1986) Binding of Clostridium botulinum neurotoxin to gangliosides, Journal of biochemistry 100, 27-33. [39] Dong, M., Richards, D. A., Goodnough, M. C., Tepp, W. H., Johnson, E. A., and Chapman, E. R. (2003) Synaptotagmins I and II mediate entry of botulinum neurotoxin B into cells, The Journal of cell biology 162, 1293-1303. [40] Dong, M., Tepp, W. H., Liu, H., Johnson, E. A., and Chapman, E. R. (2007) Mechanism of botulinum neurotoxin B and G entry into hippocampal neurons, The Journal of cell biology 179, 1511-1522.


33

Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 43

[41] Bercsenyi, K., Schmieg, N., Bryson, J. B., Wallace, M., Caccin, P., Golding, M., Zanotti, G., Greensmith, L., Nischt, R., and Schiavo, G. (2014) Tetanus toxin entry. Nidogens are therapeutic targets for the prevention of tetanus, Science 346, 1118-1123. [42] Yeh, F. L., Dong, M., Yao, J., Tepp, W. H., Lin, G., Johnson, E. A., and Chapman, E. R. (2010) SV2 mediates entry of tetanus neurotoxin into central neurons, PLoS pathogens 6, e1001207. [43] Dong, M., Liu, H., Tepp, W. H., Johnson, E. A., Janz, R., and Chapman, E. R. (2008) Glycosylated SV2A and SV2B mediate the entry of botulinum neurotoxin E into neurons, Molecular biology of the cell 19, 5226-5237. [44] Peng, L., Tepp, W. H., Johnson, E. A., and Dong, M. (2011) Botulinum neurotoxin D uses synaptic vesicle protein SV2 and gangliosides as receptors, PLoS pathogens 7, e1002008. [45] Rummel, A., Karnath, T., Henke, T., Bigalke, H., and Binz, T. (2004) Synaptotagmins I and II act as nerve cell receptors for botulinum neurotoxin G, The Journal of biological chemistry 279, 30865-30870. [46] Strotmeier, J., Willjes, G., Binz, T., and Rummel, A. (2012) Human synaptotagmin-II is not a high affinity receptor for botulinum neurotoxin B and G: increased therapeutic dosage and immunogenicity, FEBS letters 586, 310-313. [47] Rummel, A., Hafner, K., Mahrhold, S., Darashchonak, N., Holt, M., Jahn, R., Beermann, S., Karnath, T., Bigalke, H., and Binz, T. (2009) Botulinum neurotoxins C, E and F bind gangliosides via a conserved binding site prior to stimulation-dependent uptake with botulinum neurotoxin F utilising the three isoforms of SV2 as second receptor, Journal of neurochemistry 110, 1942-1954. [48] Dong, M., Yeh, F., Tepp, W. H., Dean, C., Johnson, E. A., Janz, R., and Chapman, E. R. (2006) SV2 is the protein receptor for botulinum neurotoxin A, Science 312, 592-596. [49] Tighe, A. P., and Schiavo, G. (2013) Botulinum neurotoxins: mechanism of action, Toxicon : official journal of the International Society on Toxinology 67, 87-93. [50] Rossetto, O., Pirazzini, M., and Montecucco, C. (2014) Botulinum neurotoxins: genetic, structural and mechanistic insights, Nature reviews. Microbiology 12, 535-549. [51] Montecucco, C. (1986) How do tetanus and botulinum toxins bind to neuronal membranes?, Trends in biochemical sciences 11, 314-317. [52] Hamark, C., Berntsson, R. P., Masuyer, G., Henriksson, L. M., Gustafsson, R., Stenmark, P., and Widmalm, G. (2017) Glycans Confer Specificity to the Recognition of Ganglioside Receptors by Botulinum Neurotoxin A, Journal of the American Chemical Society 139, 218-230. [53] Takamizawa, K., Iwamori, M., Kozaki, S., Sakaguchi, G., Tanaka, R., Takayama, H., and Nagai, Y. (1986) TLC immunostaining characterization of Clostridium botulinum type A neurotoxin binding to gangliosides and free fatty acids, FEBS letters 201, 229-232. [54] Yowler, B. C., Kensinger, R. D., and Schengrund, C. L. (2002) Botulinum neurotoxin A activity is dependent upon the presence of specific gangliosides in neuroblastoma cells expressing synaptotagmin I, The Journal of biological chemistry 277, 32815-32819.


34

Page 35 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

For Table of Contents Use Only

Insights into the mechanisms by which clostridial neurotoxins discriminate between gangliosides. Joshua R. Burns # †, Gregory S. Lambert # and Michael R. Baldwin *


35

Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1. HCR/A1 and HCR/A2 bind gangliosides in a similar manner. 177x177mm (300 x 300 DPI)


Page 36 of 43

Page 37 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Figure 2. HCR/A1Y1117A and HCR/A2F1117A variants enter cells more rapidly than wild type proteins. 207x544mm (300 x 300 DPI)


Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 3. HCR derivatives inhibit the activity of BoNT/A1 in primary cortical neurons. 182x186mm (300 x 300 DPI)


Page 38 of 43

Page 39 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Figure 4. Wild type HCR binding kinetics. 109x146mm (300 x 300 DPI)


Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 5. A conserved hydrophobic residue facilitates binding at the Sia5 position. 167x345mm (300 x 300 DPI)


Page 40 of 43

Page 41 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Figure 6. HCR/BI1239F binds gangliosides with greater affinity than wild-type HCR/B. 110x148mm (300 x 300 DPI)


Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 7. HCR/BI1239F binds and enters cells more efficiently than wild-type HCR/B. 163x150mm (300 x 300 DPI)


Page 42 of 43

Page 43 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Figure 8. A TeNT variant that binds specifically to gangliosides containing a Sia5 moiety. 111x75mm (300 x 300 DPI)


Insights into the Mechanisms by Which Clostridial Neurotoxins

Recommend Documents